I’ll say it again: This is NOT investment advice.
At $3 billion market cap, CIM is the second largest mREIT, but not an Agency mREIT.
In too early, Chimera was launched in November of 2007, fully nine months before Lehman Week and roughly 12-15 months until the mortgage bonds they invest in finally bottomed in price. CIM is a non-Agency mREIT, and it was set up to take advantage of the senior bonds (AAA’s) in the non-Agency MBS space trading to discounts following the pre-tremors of the meltdown (Bear Stearns mortgage hedge funds’ collapse in summer of 2007).
This group of managers does seem to lead a charmed life. When they get shellacked by the market, their friendly Wall Street dealers seem willing to go get them more money to invest.
After launching at $15 a share with 33.333 million shares, the stock went to the high teens in its first few months, but then the beginning of the meltdown came in the form of collapses of two mortgage hedge funds managed in London (Peleton Partners* and Carlyle Capital), followed by Bear Stearns. At its one-year birthday, CIM was trading under $3 a share. Still, Wall Street raised another $300 million for them in 2008 (at $2.25 per share), $500+ million in May of 2009 (at $3.22 a share), another $300+ million recently, priced under $4 a share. Only the most recent deal was accretive (priced so that proceeds were higher than the pe-secondary book value.)
CIM buys non-Agency (formerly) AAA private label MBS and re-securitizes them to create AAA bonds (called re-REMICs) to sell through the Street. By doing so, they get leverage without the risk of repo rollover. They also lever the credit exposure in the underlying bonds, but having bought the MBS at steep discounts, most of the credit risk is taken up by the discount.
In a way, the biggest risk at CIM is the timing of the repayment and/or default by the underlying mortgage borrowers, and some risk of cash flow difficulties because most of the re-REMIC bonds they issued take all the cash flows from the MBS until they get paid off.
Annaly helps out here by providing a $400 million repo credit line to Chimera, so they have friendly hands to borrow the cash to maintain their dividend, which is sitting at 16 cents a quarter on a $4 stock.
Last quarter’s published book value for CIM was $3.42 per share, so their price above $4 in today’s market translates into price/BV above 115%. To complicate the picture, that book value really can’t be verified.
It would be tough enough to value the seasoned MBS they hold, since that paper is really only half way through the resolution of the housing bubble economics. Compounding the difficulty is the fact that they did secondary deals with those MBS, typically issuing a front-end bond to investors, and holding excess interest and the remainder of the principal in portfolio.
I’ll do a non-rigorous example:
Let’s say you buy $100 million on seasoned Chase Mortgage AAA ARM MBS, with a current coupon at the floor of 4.5% for those loans. They were “prime jumbo” mortgages when originated, but origination was probably 2006 or 2007. The bonds have taken no credit losses as yet, because they are protected by the remaining AA, A, and BBB bonds, which originally amounted to 3 or 4 percent of the deal.
The good news is that you bought the bond for 60 cents on the dollar, and “only” 8% of the borrowers were 60 day’s delinquent or worse. A quick trip to the rating agencies with the updated loan tape, and you find you can issue 48 million in AAA bonds, as long as those new bonds (the re-REMICs) get every dime of cash flow until they are gone. That’s likely to be two or three years from now, with a 1.5 year average life.
The market is happy to take those 4.5% bonds off your hands at a 10% discount, so the investor is getting nearly 10% yield if the mortgages prepay and/or default as expected. (Remember, even a default results in some cash when the house gets sold, especially on those $500K to $800K jumbo mortgages on expensive houses).
So, having bought the bond for $60 million from their friendly dealer, CIM issues $48 million in bonds at 90% price, recouping $43 million of their cash. They own the other $52 million of face amount (the back end, subordinated piece) at an effective cost of $17 million, or 33 cents on the dollar.
My guess is that part of the underwriting deal with the Wall Street dealers who sell the re-REMIC bonds is up to six months of financing on that $17 million purchase (maybe for half the purchase price).
The whole underlying MBS is paying 4.5%, so that money, plus ordinary early payoffs from people who sell their houses, plus recoveries from foreclosures, plus the excess interest all combine to pay off 30% to 40% of the re-REMIC bond the first year, and even more of what’s left the second year.
All that gets CIM closer to payday. If every mortgage became a full performer, or even if losses never exceed the capacity of the AA and lower bonds to absorb, at the end of three years, roughly 16% of the principal of the re-REMIC was paid off by interest, and there’s still nearly $60 million of principal left in the Chase MBS bond. For that, CIM paid only $17 million, so even the current cash flow would amount to 15% or higher yield, along with the $60 million in principal payments.
Now, the reality. Losses are going to continue, and they will hit most likely hit even the top-priority Chase MBS bond in that bubble peak pool of mortgages. Some borrowers are going to get drastic loan mods and pay only 1% or 2% interest for now or even forever.
If the MBS bond eventually takes 2% in interest shortfall in 2011 and beyond and 30% in credit losses, the front end re-REMIC may take a year or three longer to pay off, and the bond CIM owns at the end may not have much of a coupon. The re-REMIC owners will still get their $48 million and their 4.5% interest, and CIM will have paid $17 million for a $22 million 2.5% bond that doesn’t get any cash for five or six years.
Not so nice.
Now you see why I say the book value is pretty much a mystery.
On the other hand, the 17 cents a quarter dividend is real, and likely to be stable for at least a year or two, unless the underlying mortgages get so far away from the managers’ projections that they have to recast their cash flow expectations. That gives a yield over 17%.
Unlike the amREITs, CIM will not be subject to margin calls or meaningful nightly changes in the marks on its portfolio.
We’ll see whether two years of averaging down on their investments pays off, but the picture probably won’t be clear for another two years or so.
* Peleton Partners is a classic (and drastic) example of past performance not being reliable for future expectations. They made 87% for their investors in 2007 by going short via CDS on mortgage bonds. They thought the MBS bear was done feeding, and went long. By February of 2008, they were liquidated and returned squat.